Pre-Processing and Emoji Classification of WhatsApp Chats for Sentiment Analysis

WhatsApp is among the popular social media service with over 2 billion registered users. WhatsApp is integral in people's life as a medium of communication. They use WhatsApp to share their feelings through text messages. With WhatsApp present in over 180 countries, code-switching is common. Along with this, with the increase in usage and prevalence of emojis, emojis have become indispensable during sentiment analysis. Throughout this paper, our approach to convert unstructured WhatsApp messages to a structured form is discussed on which various data mining techniques for sentiment analysis can be performed. Our approach to deal with code-mixing, different emojis and the emotions they depict, and finally, perform basic analysis using this algorithm is discussed.